What's new
What's new

Schaublin 135 spindle bearings assembly and preload procedure

Yes, kind of. You need to achieve the correct mounted preload. As bearings differ, that could mean your spacer (oil retainer..) needs to get thinner OR wider, depending on the bearing dimensions. You wont know until you start the assembly process which way you need to go.

Yes

Im not sure what they mean, that bearing should come ground to predetermined preload with a new spacer (the spacer between the two inner races is what sets this). Maybe they didnt back in the day, but you should not have to mess with that inner spacer. You can prove it by stacking it up in an arbor press and just putting a little pressure on it to see if it loads.

#7, tightening the nut wont change the relationship between the inners/outers because #7 is holding them apart

Dont know why they are having you assemble fully to check play, its a lot of work and can damage bearings. Set preload on rollers mathematically, the thrust should come "right" just tighten up. Set capture and assemble once.. Even if you have to adjust the thrust you can figure it all out before assembly.

Preload, yes.

There is a lot of bad information in this thread, and I mean no offense, but you are being lead astray by some of the comments here. In terms of spindles, this is a cakewalk, but all the fundamentals still apply, and if you dont know them its a steep and expensive learning curve.
I don’t have any past experience with spindles. Any reliable source of info from which I can learn?

What do you mean by setting the rollers bearing preload “mathematically”?
 
Last edited:
If theres a school that teaches it, Ive never heard of it. Those with this information(experience) hold it pretty close to the vest. The experience was expensive to obtain, and is tough to put in consumable pieces that someone starting from scratch can put into practice. Also many with "knowledge" have little to no practical experience. Its a bit of a niche world.

Anywho, by mathematically; The procedure you posted describes a rudimentary mathematical way of setting the preload. You can likely get away with it for this application as its super slow rpm, probably less than 4k max and youd likely not run near max much anyway. If this was in a higher speed application this equation would likely not work. There are better formulas for this adjustment, but again for your application and experience this basic one is a place to start.

All its really saying is if you assembled it all and had 8 microns of radial play, take it all back apart and adjust (shrink) the spacer (.008 x14) and you should have taken up that play. Now is that setting it to 0, or a few microns of preload, they are not clear on. Well what if you assemble the whole thing and have no radial play? Are you set prefect or do you have 20 microns of preload (interference-->to much) you have no way to know....

Another factor, say you have the thrust bearing preloaded, and the roller has no preload, you cant accurately measure the radial play since the "thrust" bearing is angular contact and will hold the shaft in both the axial AND radial directions to a significant degree. So you want to be looking at each bearing individually to set preload.

So the "right" way to do this, many will say you need a GN gauge. Most dont have one kicking around.

I attached one procedure to do this without the gauge. It takes a little doing, but this is spindle rebuilding, no one said it was easy. And obviously this assumes your parts are good, round, concentric, journals, shoulders, spacers, housings etc. You will get bad results if any one thing is not "good".
 

Attachments

  • IMG_4694.jpg
    IMG_4694.jpg
    53.2 KB · Views: 19
  • b.jpg
    b.jpg
    77.5 KB · Views: 18
"Dont know why they are having you assemble fully to check play, its a lot of work and can damage bearings."

Read the manufacturer's text. They tell you to measure the play both radial and axial *while it's assembled* to see if any changes are needed. In spec? Don't touch it. Out of spec, measure how much, dismantle it, make the changes *they* recommend, and then re-assemble. These aren't angular contact radial bearings that are ruined if you press them off a spindle throught the ball elements. It looks to be a fairly well thought-out bearing assembly.
 
"Dont know why they are having you assemble fully to check play, its a lot of work and can damage bearings."

Read the manufacturer's text. They tell you to measure the play both radial and axial *while it's assembled* to see if any changes are needed. In spec? Don't touch it. Out of spec, measure how much, dismantle it, make the changes *they* recommend, and then re-assemble. These aren't angular contact radial bearings that are ruined if you press them off a spindle throught the ball elements. It looks to be a fairly well thought-out bearing assembly.

All my points stand.

Its a bunch of work, and im not saying you will, but you could damage the bearings.
 
It looks to be a fairly well thought-out bearing assembly.
I disagree. I think it's an overly complicated design. Plenty of high quality machine tools get to the same functional result using angular contact bearings or one cylindrical roller bearing at the nose and angular contacts at the back.
 
Overly complicated, but not for the Swiss.

Heaven is where the mechanics are german, the lovers are italians, the cooks are french, the police are british, and it's organized by the swiss.
Hell is where the police are german, the cooks are british, the mechanics are french, the lovers are swiss, and it's organized by the italians.

Yes schaublin stuff is a bit fancy. But you can take this apart without damaging any of the pieces, unlike a hardinge spindle. Once you press the bearings off the spindle, throw them away. One bite at the apple.

(honesty shouldn't this be in the Schaublin sub-forum?)
 
At some point (I think 74) Schaublin decided to change the design of the 135’s spindle and got rid of the double roller bearing at the front, for a triple angular contact combo. Mine is the former design. Anyhow, this is indeed not a thread about Schaublin, but spindle bearing theory and practice in a specific case.

Aviv.
 
Meh, the design is ok, ive seen much better and much worse.

And like i said in the other thread, you can direct swap out the odd thrust bearing for angular contact thrust from a few manufacturers. But since op is trying not to replace bearings, thats not a good option.

This bearing stuff is less theory, more fact. Its just a physical thing, crush the rolling elements x amount static, it works. How you know that crush/interference spec and how to set/measure it, is where the challenges come. Engineers figured all this out decades ago obviously as very high precision bearings (and spindles) have been made for a long time, and a bearing is only as good as its mounted setup.
 
Going forward, it seems that in the lack of internal clearance gauge I have to reffer to SKF/NSK internal clearance procedures "without a gauge". The procedures are more or less similar, or at least follow the same ideas. However there is one practical question I couldn't fully figure out - how to meassure accurately the length (space) where the spacer has to be fitted. For example, ref. to SKF procedure: https://www.skf.com/us/products/super-precision-bearings/cylindrical-roller-bearings/mounting (Mounting bearings with a tapered bore by measuring radial clearance prior to mounting)-
Step #4 - "Measure the distance L between the bearing side face and the shaft abutment (fig. 9). Take measurements at different diametrical positions to check accuracy and misalignment. The difference between the single measurements should not exceed 3 to 4 μm".

In my understanding one has to measure the distance between the inner ring to the spindle shoulder, where the spacer ring will be installed for permanent clearance setting. However, I wonder what method should be used to measure this distance in absolute terms within micron accuracy as they suggest. Are block gauges the way to go?
 
Last edited:
1706642719872.png

What do you mean you cant access the location? You have the shaft sitting on a bench, slide on the roller, load it up according to the instructions, and measure the distance to the shaft shoulder. There is no access issue. If its too close for the smallest adj block, just use regular gauge block stacks til you get there.
 
"Going forward, it seems that in the lack of internal clearance gauge I have to reffer to SKF/NSK internal clearance procedures "without a gauge"

Why for heaven's sake just not follow the manufacturer's written instructions?
 
"Going forward, it seems that in the lack of internal clearance gauge I have to reffer to SKF/NSK internal clearance procedures "without a gauge"

Why for heaven's sake just not follow the manufacturer's written instructions?
Because their procedure only relevant in case the thickness spacer ring has to be reduced. Schaublin’s procedure assumes adjustment of an existing setup, i.e. always positive clearance to be eliminated. When you replace to a new bearing, who promise you the spacer ring is not too think and the bearing is over-preloaded?

I’m pretty sure that they had another factory procedure for the adjustment of new setup.
 
View attachment 425853

What do you mean you cant access the location? You have the shaft sitting on a bench, slide on the roller, load it up according to the instructions, and measure the distance to the shaft shoulder. There is no access issue. If its too close for the smallest adj block, just use regular gauge block stacks til you get there.
I assume the measurement should be taken without the front flange, so access should not be an issue IMO, but I have to take a closer look.
Are these adjustable parallels accurate (stable, finely adjustable) enough for this task when you chasing one micron?
 
It's a process I believe also requires a bit of common sense, you can throw many expensive measuring instruments at jobs like this that will only be used once in a lifetime unless you are Schaublin that is, and for the 'average Joe' like me and others its not viable so I would mix some of the guidance in manuals with some good old human intuition, eyes, feel and touch are very accurate and in many cases better than instruments which require humans to set them up!

Its important to do this right but seriously when it comes to microns you may have more chance of getting a flight in one of Elon's rockets!

You seem to have a good grasp of the task in hand so take your time, keep things clean and go for it IMO.
 
IMHO one should not measure on the bench because how tight the outer ring is in its housing also is of influence on the reading
Like I mentioned put a shim behind the existing spacer so you have positiveclearance and then measure
Then will be determened if you need to make a new spacer or can use the old one
Shim is there just to make sure you start off with a clearance on the bearing This way there is less chance of damaging the bearing

Peter
 
If you follow the bearing vendors procedure, the contribution of the interference fit is taken into account in the calculation, so measurement on a bench seems to be acceptable.

Another approach would be to follow Scahublin’s procedure. Either using a shim or additional spacing ring (however I’m not sure it’s feasible to produce such a thin super accurate spacer). Another option would be to use the existing ring, but during the assembly drive the bearing onto the taper while measuring the clearance on the bench, making sure it does not become too tight. If I’m lucky, the spacer will have to be reduced. If I reached the hard stop and the clearance is still positive, I can assemble the spindle into the housing, then continue with Schaublin’s procedure. If the hard stop was not reached and the clearance reaching zero, it means the spacer has to be extended and I have to use the other method. If I would have the right gauge blocks, I think I would start with the more generic approach (SKF/NSK). Unfortunately, I’m not properly equipped to such accurate jobs…
 








 
Back
Top